EEAT Mechanics

Stateful substrate: the layer that remembers a repeated search

Schema names show a layer that remembers what happened on earlier runs of the same query, and other repeat-query twiddlers build on it. What it stores and for how long is not known. Google confirms that past searches can shape later ones.

By Thomas Wawra· Published · Version 1.0· Systems referenced: Mustang / SuperRoot etc.

01What the schema shows

Two objects carry the name: StatefulSubstrateTwiddlerParams, the settings, and StatefulSubstrateAnnotationSignals, signals left for later steps. "Stateful" means the system keeps a memory between requests; a substrate is a base that other parts stand on. RESONEO describes it as a layer that remembers earlier runs of the same query.

The schema paths show that it is a base in the literal sense. The intent personalized boost lives in the namespace twiddler.repeat_queries.stateful_substrate, and one variant of the freshest twiddler is called FreshestSubstrate. Like every twiddler, it only recommends; the framework in Superroot, the layer that merges what Google's indexes return, decides. The parent article on twiddlers covers that framework.

Source per claim (3)
C
The stateful substrate has two schema objects: StatefulSubstrateTwiddlerParams and StatefulSubstrateAnnotationSignals.[1]
C
Other twiddlers build on it: IntentPersonalizedBoostTwiddlerParams sits under twiddler.repeat_queries.stateful_substrate, and a freshest variant is named FreshestSubstrate.[1]
C
RESONEO describes the stateful substrate as a layer that keeps state between one repeated query and the next.[2]

02What Google says about past searches

Google says openly that past searches can shape later ones. Its example is about predictions: someone searches for chocolate cake, then for "how to make", and Google may guess they mean chocolate frosting. Its help page on personalisation adds that current searches are used even when personalised results are switched off.

Both statements describe the kind of memory a stateful layer would provide. Neither names the substrate, nor says what it stores or for how long. The chocolate cake example is about autocomplete predictions, not about re-ranking results.

Source per claim (2)
O
"These predictions are based on your past searches to give you better results and help you pick up where you left off." The example concerns predictions, not result ranking.[3]
O
Even with personalised recommendations turned off, Google says it still uses location, language, device type and current searches.[4]

03E-E-A-T

The stateful substrate is plumbing for personalisation and has no E-E-A-T bearing. Whether it holds state for a session, for a signed-in account or both is unknown. It explains why the same query can return a different order the second time, and why a single rank check says little about how a page ranks for everyone.

Source per claim (1)
C
The stateful substrate has no E-E-A-T bearing; whether it keeps state per session, per account or both cannot be read from the schema.[5]

04References

This Deep Dive is Layer 2 content. It interprets the material and always points back to Layer 1, the reference layer. Every claim is mapped to a source, labeled by type: [A] Sworn testimony · [B] Verbatim from the leak · [O] Documented by Google · [W] Peer-reviewed study · [S] Checked on the page · [C] Our assessment · [P] Patent.

© Thomas Wawra · Senior SEO Manager · wetter.com — a Funke Digital company